This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [PATCH, middle-end]: Fix PR tree-optimization/{31966, 32533}

On 7/17/07, Mark Mitchell <mark@codesourcery.com> wrote:

>> 2007-07-16  Richard Guenther  <rguenther@suse.de>
>>             Uros Bizjak  <ubizjak@gmail.com>
>>
>>         * tree-if-conv.c (find_phi_replacement_condition): Unshare
>> "*cond"
>>         before forcing it to gimple operand.

I don't fully understand the consequences of this proposed patch for
4.2.  Without checking enabled, do we have a real problem without this
patch?  In other words, what problem is solved by the patch, for users
building with checking disabled?  (These are not rhetorical questions; I
just don't understand fully.)


The compilation of the testcase form the PR fails with "invalid tree
node sharing" when checking is enabled, otherwise compiles OK. There
is another C runtime testcase that fails without master patch, and
there is no "invalid tree node sharing" even with checking is enabled.

Uros.

Thanks,

--
Mark Mitchell
CodeSourcery
mark@codesourcery.com
(650) 331-3385 x713


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]